Ten years ago, the world spent more building fossil fuel supply than building electricity systems. A decade on, that relationship has inverted completely. “Now, it is exactly the opposite,” said Dr Fatih Birol, Executive Director, International Energy Agency (IEA), describing the reversal at the agency’s World Energy Investment 2025 launch.

Global energy investment reached USD 3.3 trillion in 2025, with USD 2.2 trillion flowing to clean energy against USD 1.1 trillion to fossil fuels. That two-to-one ratio has held for at least three consecutive years, according to IEA data.

For institutional investors with Asia energy exposure, the aggregate number conceals the more useful story. Inside the region, that capital has split into two entirely different trades, and treating them as one is starting to cost money.

The Two Speeds Inside One Region

Japan and Korea now send 92% of their energy investment to clean technologies. That compares with a global average of 66%, according to the IEA’s most recent country-level data. Both economies import almost all their fossil fuels.

That is why the incentive to build clean capacity arrived earlier there than almost anywhere else in Asia.

Competition in these markets is intense, permitting is established and returns have compressed accordingly. What remains attractive is the layer beneath generation: grid hardening, transmission and the storage and services infrastructure that keeps an already-built system resilient.

Southeast Asia is running a different race entirely. Fossil fuel investment in the region fell from USD 70 billion in 2015 to USD 50 billion in 2025, a 29% decline. Clean energy investment rose from USD 30 billion to USD 47 billion over the same decade, up 57%.

That is not a mature market maturing further. It is a growth market still being built, and the two should not sit in the same allocation model.

Where the Real Opportunity Sits

Indonesia illustrates the gap most starkly. The country’s Energy and Mineral Resources Ministry puts its renewable technical potential at 3,687 gigawatts. Installed renewable capacity, as of June 2026, stood at just 16.32 gigawatts, a little over 0.4% of what the resource base could theoretically support.

None of this technical potential converts into a portfolio position on its own. A gap this size, between what the resource base could support and what is actually built, cannot close from one balance sheet.

Indonesia’s state utility, PLN, cannot fund it alone. That is precisely why infrastructure funds and family offices are underwriting projects directly, rather than waiting for state capital to close it.

India shows what closing that gap looks like in practice. “We are on track,” said Pralhad Joshi, India’s Minister for New and Renewable Energy. His Ministry confirmed 283.46 gigawatts of total non-fossil capacity installed as of 31 March 2026, already past the halfway mark toward the country’s 500-gigawatt 2030 target.

By this month, that milestone had advanced further still. Non-fossil capacity climbed to 300.5 gigawatts as of 31 July 2026 – up from the 283.46 gigawatts recorded four months earlier – crossing 60% of the 2030 target. India added a record 44.6 gigawatts of solar and 6 gigawatts of wind in the last financial year alone.

The pattern holds across the region’s import-dependent economies.

Every gigawatt of domestic renewable capacity is not only a climate commitment. It is a subtraction from a fuel import bill that currency and fiscal shocks have made harder to ignore since the Strait of Hormuz closure.

The Diligence Question Has Changed

Growth-market infrastructure in Asia has long been underwritten as a political risk story: unpredictable regulation, inconsistent governance, uncertain contract enforcement.

That framework has not kept pace with what has actually changed on the ground.

Countries that need reliable domestic power to protect their currencies and fiscal positions have a structural incentive to honour long-term energy contracts, not break them. The risk has not disappeared. It has moved.

“The nature of risk is shifting from contract security to execution,” said Rahul Agrawal, Managing Director, Energy Infrastructure, Actis, in a signed column published 10 August 2026.  Grid readiness, permitting speed and the reliability of domestic supply chains now matter more than the legal protections a term sheet can offer.

That is a different due-diligence exercise than the one most infrastructure funds built their models around. It rewards investors who can assess whether a project will actually get built on schedule, not just whether the contract looks defensible on paper.

Not every part of that story is settled. Grid interconnection across Southeast Asia depends on member states delivering shared infrastructure on a common timeline. Coordinating 10 distinct regulatory systems is a genuinely harder problem than any single country’s permitting reform.

Execution risk has not been eliminated. It has simply replaced a different kind of risk that investors were already pricing.

What This Means for Capital Deployment

The two trades require different underwriting entirely. One is a yield and resilience play, competing on grid and storage assets where the generation build-out is largely finished.

The other is a construction and execution play, where the return depends on delivery capability, not on the strength of the paperwork.

A single Asia energy allocation priced off one side will misread the other every time. The investors who split their underwriting now, before the next allocation cycle, will be pricing the region that actually exists. Their models still assume the old one.

SK Hynix just posted the best quarter in its history. Investors sold it anyway.

Operating profit surged 557% year-on-year to KRW 60.54 trillion (approximately USD 41.6 billion), an all-time record, on 29 July 2026. The stock closed 9.61% lower the same day, because the record still landed short of the roughly KRW 64 trillion analysts had priced in.

Thirteen days earlier, TSMC had made the same bet, only louder. It raised its 2026 capital expenditure guidance from USD 52–56 billion to USD 60–64 billion. “Our conviction in the multi-year AI megatrend remains very high,” said C.C. Wei, Chairman and CEO, TSMC, 16 July 2026.

TSMC’s US-listed shares fell anyway.

Two of Asia’s most important chip companies had just proved the same thing inside a fortnight: a record number is no longer enough. What “beating expectations” requires has quietly changed and it did not stop at chipmakers.

Beating Earnings Stopped Being Enough

SK Hynix and TSMC were not isolated cases. In the same week SK Hynix reported, 20 of the world’s most valuable chip stocks lost USD 1.3 trillion in combined market value. The figure comes from a CNBC analysis of FactSet data. Nvidia lost USD 238 billion. SK Hynix lost USD 176 billion, Samsung Electronics USD 173 billion, TSMC USD 119 billion, Micron USD 113 billion and AMD roughly USD 110 billion.

Michael Field, Chief Equity Strategist, Morningstar, told CNBC on 29 July 2026 that the moves reflect sentiment rather than fundamentals. “Simply put, it’s loss of confidence,” he said. Charlie Dai, VP Principal Analyst at Forrester Research, pointed to a different worry: that AI infrastructure spending had reached its high point sooner than the market expected.

The Demand Side Has Its Own Cash-Flow Problem

The companies buying the chips face the same scrutiny as the companies making them. Alphabet’s free cash flow turned negative for the first time since its 2004 initial public offering, at USD 5.9 billion for the second quarter of 2026. Capital expenditure had doubled year-on-year to USD 44.9 billion. Management raised full-year 2026 capex guidance to USD 195-205 billion, up from USD 180-190 billion.

Meta’s quarter told the same story with sharper numbers. Free cash flow fell 91% year-on-year, from USD 8.55 billion to USD 784 million. Quarterly capital expenditure reached USD 31.1 billion. Meta narrowed its full-year 2026 capex guidance to USD 130-145 billion by raising the floor, not the ceiling.

Both companies beat revenue estimates. Both were punished anyway, on the demand side of the same trade SK Hynix and TSMC had already shown on the supply side.

What the Boardroom Already Knew

None of this should surprise anyone who read the research published months before the sell-off. KPMG’s inaugural Global AI Pulse survey drew on more than 2,100 executives across 20 countries. It found that Asia-Pacific firms plan to invest an average of USD 245 million in AI over the next 12 months. That is the highest of any region, well above the USD 186 million global average.

IDC forecasts that by the end of 2026, 45% of AI-fuelled digital use cases across Asia-Pacific and Japan will fail to meet their return-on-investment targets. IDC attributes the shortfall to poor data foundations and unclear value realisation, not the technology itself.

Gartner named the dominant frame for the year in January 2026. “AI is in the Trough of Disillusionment throughout 2026,” said John-David Lovelock, Distinguished VP Analyst, Gartner. Its most recent forecast, published in May 2026, put worldwide AI spending at USD 2.59 trillion for the year, up 47% year-on-year.

Grant Thornton’s 2026 AI Impact Survey covered nearly 1,000 senior US business leaders. It found that 78% lack confidence their organisation could pass an independent AI governance audit within 90 days. That figure is US-only, but the mechanism it names travels. Three in four boards approved major AI investment, yet only 52% set clear governance expectations for it.

Two Symptoms, One Diagnosis

The market is not pricing AI’s technology risk. It is pricing a monetisation-proof problem that KPMG, IDC and Grant Thornton had already quantified before the chip sell-off began. Investors are now reacting, in real time, to a gap that internal CIO and AI Impact surveys were already flagging months earlier.

Enterprises are spending at record levels and cannot yet reliably show the spend is converting into value. Asia-Pacific sits at the centre of both halves of that trade. It is the region with the highest planned AI spend and the one IDC forecasts will see the highest rate of ROI failure.

Not every analyst reads the sell-off as a verdict. Bank of America’s Vivek Arya has long pushed back on that logic. Markets cannot simultaneously price AI capex as unsustainable and AI adoption as powerful enough to make software obsolete, he argues. The two premises, in his view, cannot both be true.

That argument has force and it complicates any simple bear case on AI infrastructure. It does not, however, address what KPMG, IDC and Grant Thornton had already measured directly. The question was never whether AI works. It was whether enterprises spending on it can prove that it does.

What Closes the Gap

The proof gap does not close by next quarter’s earnings call. It closes when Asia-Pacific enterprises can show, not claim, that AI spending converts into measurable value. That is a governance problem, not a technology one.

Until then, every capex headline from here to Q4 will be read against a single number. IDC put it at 45% of Asia-Pacific’s AI use cases missing their return this year, published months before a single chip stock fell.

Earnings season no longer prices spending on faith. It prices spending against a number the boardroom already knew.

Vietnam attracted USD 38.06 billion in registered foreign direct investment in the first seven months of 2026, up 58% year-on-year, the National Statistics Office reported. Manufacturing and processing took USD 11.58 billion of the USD 21.05 billion in newly registered capital across 2,429 projects. Singapore led all investors at USD 7.5 billion, ahead of South Korea, Hong Kong and China. Disbursed FDI reached USD 15.2 billion, the highest seven-month total in five years.

Editor’s View: Singapore supplies over a third of Vietnam’s new FDI – a concentration that flatters the headline number but narrows the base. When one source accounts for more capital than the next three combined, “record inflows” is really a bet on one relationship holding.

Read full story →

When the UN Office on Drugs and Crime launched its latest threat assessment in Bangkok on 21 July 2026, it reframed Southeast Asia’s organised crime economy. Rather than a collection of national criminal ecosystems, the report describes a franchised network. Its operators share laundering channels, trafficking routes and digital infrastructure across borders.

“Their operating model looks like corporate franchising,” said Delphine Schantz, UNODC’s Regional Representative for Southeast Asia and the Pacific. Laundering, trafficking and data-harvesting run as specialised departments plugged into one shared service network.

For banks, that finding lands awkwardly on top of a screening framework built for a different kind of criminal economy. If organised crime increasingly operates as a cross-border network rather than a collection of national ecosystems, the appropriate unit of financial crime risk is the network. It is not the country it happens to touch.

The jurisdictions most associated with this network currently sit on three different rungs of the FATF risk ladder. One is blacklisted, one is grey-listed, one carries no FATF designation at all. UNODC says the same operators move between them regardless.

The Departments Inside One Franchise

UNODC’s report, An Interconnected Criminal Ecosystem: Transnational Organized Crime Threat Assessment for South-East Asia 2026, documents a shift.

Groups that once confined themselves to a single territory or a single type of crime now run several at once. Scam operations, human trafficking, migrant smuggling, drug trafficking and money laundering all move through shared financial and technical infrastructure.

Larger operators sell laundering pipelines, recruitment funnels and fraud platforms to smaller affiliates for a fee. UNODC describes this as a “crime-as-a-service” model, with specialist providers offering money laundering, recruitment and digital infrastructure to affiliated groups.

That service model extends even to purchased access inside corrupt police forces, immigration departments and corporations.

Losses across East Asia, Southeast Asia, Australia and New Zealand rose from an estimated USD 18–37 billion in 2023 to USD 88.3–114.1 billion in 2025.

UNODC lead analyst Inshik Sim said the network’s scale is “outpacing existing responses,” designed for far less sophisticated criminal activity than the region now faces.

One Network, Three Jurisdictions

Jurisdictional assessments remain indispensable for measuring the strength of a country’s AML and supervisory framework. The challenge is that organised criminal networks increasingly operate across those jurisdictional boundaries rather than within them.

Myanmar sits on FATF’s blacklist, formally the list of High-Risk Jurisdictions Subject to a Call for Action. FATF calls for enhanced due diligence proportionate to the risk there.

Laos sits on the grey list, added in February 2025 and still under increased monitoring as of FATF’s 19 June 2026 plenary statement. Cambodia, where the network’s most heavily sanctioned entities are domiciled, carries no FATF designation at all.

A bank’s jurisdiction-based screening model treats these three inputs completely differently by design. Maximum scrutiny applies to Myanmar, enhanced monitoring to Laos, and standard due diligence to Cambodia.

The Cost of Treating Networks as Countries

For financial institutions, that structural mismatch carries a measurable cost.

LexisNexis Risk Solutions surveyed the region in 2023, publishing its findings in March 2024. It put Asia-Pacific’s annual financial-crime compliance bill at roughly USD 45 billion across five markets, with 98% of institutions reporting costs still rising.

Although based on a 2023 survey published in 2024, it remains one of the few publicly available estimates of financial crime compliance costs in Asia-Pacific. Much of that investment still supports control frameworks built around customer, transaction and jurisdictional risk. That same network operates across those boundaries by design.

The Bank for International Settlements has documented the long-term decline in correspondent banking relationships, particularly affecting smaller and higher-risk jurisdictions. That trend reflects the growing cost of managing financial crime risk across borders.

FATF’s own guidance on correspondent banking encourages institutions to adopt genuinely risk-based assessments rather than relying solely on jurisdictional indicators. The guidance reflects an acknowledgement that country ratings alone cannot capture every source of financial crime risk.

When Regulators Target the Network Instead

One action against this network actually worked, because it targeted the network, not a country.

FinCEN first proposed severing Cambodia-based Huione Group from the US financial system under Section 311 of the USA PATRIOT Act in May 2025. It finalised the rule on 14 October 2025, alongside coordinated OFAC and UK sanctions that hit 146 targets inside Prince Group.

The finalised rule found Huione had laundered more than USD 4 billion in illicit proceeds between August 2021 and January 2025. Those funds moved for actors spanning Southeast Asian scam operations and North Korean cyber theft alike.

Treasury widened the sanctions against the same network again in June 2026.

“Scam centers in Southeast Asia steal billions of dollars from American victims each year,” said Scott Bessent, Secretary of the Treasury, in the accompanying statement. Each action treated the network as the unit of analysis, not the jurisdiction it happened to be sitting in that month.

How Regulators Are Redefining the Unit of Risk

Singapore’s regulator got there by a different road entirely.

On 4 May 2026, the Monetary Authority of Singapore began a Proof-of-Value exercise, pooling transaction data from five banks. It marked the first time MAS has combined bank-level data at the regulator tier, built for pre-emptive detection rather than post-loss reporting.

The rationale MAS gave is structural, not political. Any single institution only holds transaction history for its own customers. That limits how much of a fraud pattern spanning several banks can ever be visible to any one of them alone.

That is precisely the blind spot a franchised, service-sharing criminal network is built to exploit.

The Question This Leaves Banks

UNODC did not set out to write a banking compliance paper, but Schantz’s own language answers the question anyway. A network organised like a franchise is not disrupted by removing one franchisee. It is not seen by a control built to flag one jurisdiction at a time.

MAS’s pooled-data pilot and the Section 311 action against Huione are early proof of a shift already underway. Jurisdiction-based screening is giving way to network-based screening, for the institutions paying attention.

Every other bank still pricing risk by border will keep paying for the gap in its own model. That is the true cost of mistaking one franchise for three unrelated countries.